Measuring Subjective Dimensions of Empowerment Among Extremely and Moderately Poor Women in Colombia and Peru. Lessons from the Field

Transcription

1 Measuring Subjective Dimensions of Empowerment Among Extremely and Moderately Poor Women in Colombia and Peru Lessons from the Field Susana Martinez-Restrepo, Fedesarrollo Johanna Yancari, Instituto de Estudios Peruanos Laura Ramos Jaimes, Fedesarrollo Women face economic, social, and cultural challenges that limit their access to markets, quality jobs, and entrepreneurship. Women s economic empowerment is gaining interest among academics, international organizations, practitioners, and policy makers. The way women s economic empowerment is defined and measured impacts the design of interventions, programs, and policies. More accurate definitions and measurements will enhance the impact of policies and programs on women s economic empowerment. The challenge is that we still do not know the best ways of measuring women s economic empowerment. Traditional measures use employment, income, and education as proxies for economic empowerment. More recently there has been recognition of the relevance of subjective dimensions to measure women s economic empowerment, such as decision-making power over purchases, bargaining power, subjective perceptions of well-being, and freedom of choice. In this think piece, we discuss our experience implementing quantitative and qualitative instruments, including subjective measures, to assess economic empowerment among poor women in Colombia and Peru. We conclude that in these specific cases, the instruments used to measure empowerment through subjective dimensions do not work for poor women. The difficulty these women have in understanding abstract concepts, their cultural definitions of decision making showed us that researchers should adopt different strategies when measuring subjective empowerment among poor women. This think piece is part of a larger research project aimed at comparing the pros and cons of existing methodologies to measure women s economic empowerment in Latin America, Asia, and Africa. 1 Measuring women s economic empowerment Empowerment is often defined as the increasing capacity of individuals or groups to transform choices into actions, to acquire the necessary capabilities to achieve autonomy, and to make choices and exert influence over decisions that affect their lives (Petesch et al., 2005; Molyneux, 2008; Kabeer, 2005). Kabeer (2001) suggests that this definition is particularly applicable in contexts where this ability has been denied to individuals. Women s empowerment is related not only to their ability to make choices, but also to their ability to control resources that give them power within and outside of the household (Sen, 1999). According to authors such as Agarwal (1997) and Sen (1981), women s access to economic assets and their ability to survive outside of 1 This study is being developed with the support of Canada s International Development Research Centre (IDRC) and the program Growth and Economic Opportunities for Women (GrOW). 1

2 the household determine the equality or inequality of their bargaining power. Thus, there is general agreement that integration of women into the labor market is a key indicator of women s intra-household bargaining power (Sen, 1999; Kabeer, 2005; Kabeer et al., 2011). Typically, variables such as education, employment, and income have been used across different contexts to measure economic empowerment (Ibrahim and Alkire, 2007; Samman and Santos, 2009). Other critics have suggested that these variables fall short when we view empowerment as an expansion of agency. These more traditional variables do not always capture aspects of empowerment associated with agency. By definition, agency is subjective. It entails a process by which an individual evaluates his or her goals and undertakes action to change an object or a circumstance in a given context or environment. Sen defines agency as what a person is free to do and achieve in pursuit of whatever goals or values he or she regards as important. (Sen, 1985, p. 206). So, in addition to considering such variables as education and income as proxies for empowerment, we need to measure what it takes for an individual to act. Some researchers have gone beyond economic measures of empowerment to propose notions of agency such as autonomy, self-determination, liberation, political participation, mobilization, and selfconfidence (Narayan, 2005). Ibrahim and Alkire (2005), Alkire et al (2013), Narayan et al. (2000) and the World Bank (2001), and more recently (United Nations Foundation and ExxonMobil Foundation, 2015), have proposed different ways to measure women s empowerment using subjective dimensions, including the ability to negotiate or bargain around household economic decisions, particularly regarding household expenditures. Diener and Biswas-Diener (2005) recommend measures of self-efficacy, self-confidence, mastery, and communal efficacy. The report by United Nations Foundation and ExxonMobil Foundation (2015) introduced and discussed measures of satisfaction with life (overall satisfaction and stress level), self-confidence (overall selfconfidence and willingness to assert oneself and to take risk), and self-esteem. Recent studies have done important work compiling instruments used to measure women s empowerment (Knowles, 2015). These include subjective measurements related to the labor market (such as labor trajectories, meaning of work, motivations), as well as other subjective dimensions (such as self-esteem, self-confidence, gender roles and norms, and life satisfaction). But how do we know these different measurements work to capture agency, self-esteem, and autonomy as key aspects of empowerment? Frequently, surveys designed as part of impact evaluations of interventions have found no effect on these variables. However, when qualitative studies accompany these impact evaluations, they do tend to show an impact on these subjective dimensions of empowerment. So, did the program have no impact (as measured through quantitative indicators) or, given that qualitative studies show an impact, did researchers get it wrong when defining measures for subjective dimensions of empowerment? Below we discuss some lessons learned in Colombia and Peru from applying questionnaires that aim to measure subjective dimensions of women s empowerment. 2

3 1. Is decision-making power over household purchases really a good proxy for women s empowerment? The literature has shown during the last two decades that decision making and intra-household bargaining are key subjective proxies for women s empowerment. Agarwal (1997) and Duflo (2012) suggest that the proportion of income family members contribute to the household can have a significant impact on how much each household partner influences decisions regarding the purchase of goods and services. Thus, giving women cash benefits could increase their bargaining power in the household s decision-making process, particularly regarding consumption decisions. Furthermore, in 2004, Sanders and Schnabel showed that the use of savings accounts among abused women with limited resources leads to financial independence, a necessary prerequisite for leaving their abusive partners. Karlan et al. (2007), found through an experiment conducted among women in the Philippines that the use of financial products has a significant impact in empowering women by increasing their capacity to make decisions within the household. Based on this evidence, we conducted an impact evaluation of the Pilot Savings Program (PSP) among beneficiary households of the Conditional Cash Transfer Program (CCTP) JUNTOS in Peru. We aimed to identify the effects of strengthening women s financial capacities and savings on gender relations within the household. 2 We measured these effects through the decisionmaking power women wield over expenditures in their households. We asked, for example: who makes the decisions about minor purchases such as school uniforms or groceries and about major expenses like buying a TV or repairing the house? Possible responses included: women alone, partners alone, both women and their partners, or women and other family members. This distinction is important since women are more likely to make decisions over minor household purchases than over major purchases that require more negotiation, money and sometimes even loans. We assumed that women participate in decision making when they make the decisions alone or together with their partners or another family member. We also assumed that if women are more likely to save, they are also more likely to participate in major household purchase decisions. If this were the case, we could infer that women had greater decision-making power within the household and were more empowered. Our results showed that women s participation in household decision making about expenses was surprisingly high. More than 96% of women in the baseline (in control and treatment) mentioned that they participate in decisions for minor expenses. This result increased to more than 98% in the exit poll. We saw a similar result regarding major expenses, where more than 88% of women mentioned in the baseline that they participate in these decisions, and in the exit poll this percentage rose to more than 93% (in both control and treatment). 2 Between 2010 (baseline) and 2012 (exit poll), the Instituto de Estudios Peruanos (IEP) conducted an assessment of the Pilot Savings Program (PSP) among beneficiary households of the Conditional Cash Transfer Program (CCTP) JUNTOS in Peru. The CCTP JUNTOS is a program that targets the population in poverty and located in rural areas. The PSP aimed to promote and encourage savings among the users of the CCTP JUNTOS through training lectures on financial education during a sixmonth period. The hypothesis, based on the gender and financial education literature, was that women who have a savings account are more empowered financially, which is considered a first step toward economic empowerment (Johnson and Sherraden 2007). 3

4 Table 1. Decision making at home: minor and major expenses Baseline Exit Poll Differences Variable Treatment Control Treatment Control in Differences Women participate in decisions on minor household expenses 97,8% 96,2% 99,4% 98,1% -0,3% Women participate in decisions on major household expenses 89,0% 88,6% 93,2% 93,1% -0,3% Observations Source: Baseline and Exit Poll Surveys of PSP among beneficiary households of the CCTP JUNTOS These results run counter to those obtained from in-depth interviews with a group of ten women who received training in financial education. The cultural characteristics of these indigenous women from poor rural areas influenced the results. The group of ten women interviewed indicated that they participated in decisions about major expenses because they considered that communication from their husbands on the spending decision included them, even if there had been no discussion of the actual expenses. We argue that decision-making power over purchases is not necessarily a proxy for empowerment, particularly in the case of poor Latin American women. An alternative explanation for the findings is that decision making power is important for women's empowerment, but that it was not measured well in the large study. Qualitative methods can give us new ideas on how to create better measures regarding household decision making. In the case of Peru, we can attribute the results indicating high levels of engagement in decisionmaking among women to cultural definitions of decision. As the qualitative evidence revealed, women did not participate in making major household decisions. Their understanding of participation in decision making is based on a patriarchal culture in which men traditionally make decisions and women agree to and implement these decisions. These women consider the simple fact that men communicate their decisions to women to be a democratic practice. The high percentage of women who exercised control over minor expenses could be related to the place given to women as super mothers in Latin American societies. Indeed, this concept of super mothers, developed by Folbre (1994), suggests that women have to continually make decisions about small purchases (e.g., groceries and child-related expenses) to ensure correct home management. This situation, in combination with the rural condition of this population, in which men usually migrate seasonally to find work outside their communities, leaves the entire responsibility for the household to women. We suggest that this decision-making control is evidence of women trying to be super mothers by doing it all, a situation that creates time constrains that could affect other dimensions of women s economic empowerment such as labor market participation. 2. The challenges to understanding the concrete and the abstract within subjective measures of empowerment: the case of extremely and moderately poor women in Colombia and Peru Another way of looking at women s empowerment is through subjective well-being and its relationship to psychological empowerment. Subjective well-being refers to the evaluation 4

5 people make of their lives and includes pleasant emotions, fulfilment, and life satisfaction (Diener and Biswas-Diener, 2002). Psychological empowerment is one facet of subjective wellbeing since it involves people s beliefs about whether they have the resources, energy, competence, and freedom to accomplish their goals (Diener and Biswas-Diener, 2002) to attain the life they value (Sen, 1999; Sen, 1985). In this section, we discuss our experience after implementing a pilot questionnaire about subjective dimensions of empowerment among women under extreme poverty, victims of violence in Colombia 3, and rural women living in poverty in Peru. Based on Sen (1985, 1999), Ibrahim and Alkire (2007), and Samman and Santos (2009), we piloted some modules to capture subjective dimensions of well-being, self-esteem, and empowerment and also tested the widely used staircase of freedom. In Colombia, a total of 83 women answered the pilot questionnaire, which was distributed in rural and urban areas the states of Antioquia, Valle del Cauca and Bogotá D.C. On average, participants were 40 years old, and 75% had less than a high school diploma. Almost half of the participants (44%) reported that they had worked during the previous week. The other half (46%) reported they dedicated most of their time to unpaid household chores. Average reported income was 72 USD per month and 9.23% reported not receiving any type of income. The Peruvian case included poor women in rural areas as discussed in the first section. Based on Sen s definition of agency and empowerment (1999), the objective of the pilot was to understand what was really important for women living under extreme poverty without making any assumptions about what mattered in their lives (for example, the preconceived idea that material well-being is important). Furthermore, we wanted to know how free they felt to make choices about the aspects of life they valued and the goals they wanted to achieve. As Figure 1 shows, 84% of women responded affirmatively to the question Would you like to change something in your life? We expected the number would have been greater, taking into account the economic conditions of all women in the sample. This response would mean that 16% of women living in extreme poverty, often in unhealthy sanitary conditions and lacking adequate food, do not want to change anything in their lives. The second question was open-ended What aspects of your life would you like to change? and gave women the option to report the three aspects most important to them. The most common responses included Change my labor situation, Solve personal problems, Become 3 This pilot was part of the qualitative component of the impact evaluation of Red Unidos, the Colombian strategy to eradicate extreme poverty. The Red UNIDOS strategy provides psychosocial support for families through a social worker who visits between two and eight times per year. Although women s empowerment was not a program goal, it was expected that through the work of social workers and the achievement of these goals, women would gain further economic empowerment and intrahousehold bargaining and decision-making power. Several Red Unidos impact evaluations showed that the program did not manage to reduce extreme poverty or empower women beneficiaries. A qualitative study followed in 2013 that aimed to explain the lack of significant results and to get deeper insights about the program s lack of effectiveness. For more information about the studies, see: Fedesarrollo, Econometría, SEI and IFS (2012), Evaluación de Impacto de Juntos (hoy Unidos). Red de Protección Social para la Superación de la Pobreza Extrema Informe de Evaluación: Diciembre de 2011; Technical Report; Abramovsky, Laura; Attanasio, Orazio (2016). Challenges to promoting social inclusion of the extreme poor: evidence from a large scale experiment in Colombia. 5

6 a home owner, Get more education or training, and Make home improvements. All these were concrete. To the question Who or what would help you to achieve that change? the most common response was myself, followed by God or Jesus Christ, my family, and finally, the national government, and the local government. In this case, we provided the options and women chose the three that best represented their beliefs. While myself remained the most common first response, it is interesting to note how important Jesus Christ and the family is to help them improve their life. Leaving change to Jesus, the family, might take away the sense of agency women need to make the necessary changes to improve their lives and attain their goals. Leaving change to the national and local governments is more difficult to understand. On the one hand it can show certain dependence to the State s subsidies to change ones lives. On the other hand, it can show that women are demanding better social policies. Figure 1. The life we value 1. Would you like to change something in your life? Yes (84%) No (16%) 2. What aspects of your life would you like to change? (Open question: three most important aspects) (1) Change my labor situation (54,29%) (2) Solve personal problems (27,14) (3) Become a home owner (27,14%) (4) Get more education and trainning 25,71%) (5) Make home improvements (24,29%) (1) Myself (73,13%) 3. Who or what would help you to achieve that change? (2) God or Jesus (58,21%) (3) My family (31,34%) (4) The National Government (25,37%) (5) The Local Government (19,4%) 4. Do you have the freedom to chose or to decide about those changes? 47% (10th staircase) today and 18% (10th staircase five years ago) Source: Authors elaboration based on the questionnaire s structure and descriptive language. Finally, we included the following commonly used question: Imagine a staircase of ten steps, where at the bottom, on the first step, people cannot choose or make decisions regarding their lives and on the top step, the tenth, people have complete freedom to choose or make decisions regarding their lives. On what step of the decision-making staircase are you? We created a template to identify which step the woman placed herself on at the time of the survey and to compare that placement with her position on the staircase ten years earlier (see picture in Figure 2). Data from Colombia showed that 47% placed themselves on the tenth step (the highest), in terms of their perceived freedom to choose or make decisions about changes. For the most part, respondents perceived they had more freedom today than ten years ago. We noticed, however, 6

7 that women misinterpreted the question and were able to respond only after the interviewers explained the question many times. The situation was similar in Peru (see below) where it took, on average, between five and eight minutes for pollsters to explain the question. The case of Maria illustrates the difficulty women had in understanding this question (see Figure 2). Figure 2. The staircase Interviewer: Imagine you are on a staircase. The first step represents no freedom at all to decide about your life and what you want, while the tenth step represents total freedom to do so. On which step are you standing right now? (The interviewer shows the card with the stair case). Maria: Now I live with my husband and my children. Interviewer: On which step were you standing five years ago? Maria: Five years ago we were living with my mother-in-law. Like Maria s responses, many of the women s answers were not truly related to the question asked. On the one hand, Maria s direct response was based not on her current freedom to decide over her life, but on the location of her residence today and five years ago. On the other hand, it is also possible that this response indicated that she is doing better today than five years ago, or that she has more freedom today since she is living with her family in her own household. She may have used a concrete example to convey the abstract concept of freedom. This does not mean that freedom is not important to her but that might be expressed differently. A good research questionnaire should allow abstract concepts to be applied to Maria s concrete context. We observed that all the responses had something in common: they were concrete and associated a point in time with a physical place. The question used the staircase as a metaphor, combining a concrete concept (a staircase) with an abstract concept (freedom to choice and freedom to decide). Because of this, we suspected that the answers obtained after the interviewers numerous explanations might be biased by the interviewers own opinions and perceptions. This outcome is consistent with evidence found in Peru in the same impact evaluation described in the first section. As shown in the table below, the differences between treatment and control in responses to this question were not statistically significant. In both cases, women in the treatment and control group showed the same results in both the baseline and exit polls. The field work showed that the women surveyed did not understand the question. For example, the average time spent on this question according to records was about eight minutes, due to the pollster s numerous explanations. The answers to this question in most cases came from the interpretation of the pollster, rather than the response of the interviewee, and in the end the researchers decided not to use the results. Table 2. Staircase regarding decision making: At present, 5 years ago and 10 years ago 7

8 Baseline Exit Poll 1/ Difference Variable Treatment Control Treatment Control in Difference Staircase regarding decision making 5 years ago (1-10) 5,22 5,09 6,66 6,50 0,04 Staircase regarding decision making 10 years ago (1-10) 4,11 4,22 6,66 6,50 0,28 Observations / The results of the exit poll answer the question: At present, what step of the staircase regarding decision making are you? Source: Baseline and Exit Poll Surveys of PSP among beneficiary households of the CCTP JUNTOS. Evidence from this section shows that extremely poor women might have difficulties understanding abstract concepts evaluated on surveys. They express and understand their goals and the things they value in their lives in terms of concrete objects or pursuits, such as holding a job, having a house, repairing the house, or getting more training. To accurately understand the life women value, and therefore the process of agency and empowerment, questionnaires and qualitative methods need to include concrete objects, actions or situations (e.g., being a homeowner, getting a job, making home improvements). In behavioral experiments, economists Mullainathan and Shafir (2013) showed that a life of scarcity reduces the mental bandwidth available to deal with the problems such scarcity generates. This means, for example, that many individuals that have lived under scarcity do not necessarily consider other objectives or needs that are not necessarily linked to the correspondent scarcity or to their concrete needs that are relevent to their life. Strategies such as extra training of pollsters and visual aid cards might help, but understanding subjective dimensions might be challenging for women who can barely read or understand a text when they do manage to read. Questionnaires measuring subjective dimensions of empowerment among women in poverty must make an effort to transform the abstract into the concrete according to their specific context, education level or human capital. Finally, it is crucial to understand the local context of women and their cultural definition of empowerment when designing implementing thess types of instruments. 3. Final remarks and methodological implications This think piece discussed the challenges that arise when trying to measure different dimensions of empowerment among poor women in Latin America, with a focus on subjective dimensions. We want to highlight three methodological implications from our experience. First, women living in poverty with low levels of education do not clearly understand subjective measures of well-being, including abstract concepts (freedom, freedom of choice, and freedom in decision making for major and minor purchases). Second, we suggest that these instruments do not necessarily work for women experiencing poverty in Latin America. Third, we show that decision-making power over household purchases is not necessarily a good proxy for empowerment. We also discussed the difficulty we had applying questionnaires containing measures of subjective well-being and using the staircase as a way for women to define their degree of freedom to choose. Because they defined their goals and values in terms of concrete objects 8

9 (having a house, repairing the house, getting more training) women found it difficult to understand abstract questions around values. We conclude that due to the difficulty women had understanding abstract ideas, these instruments should not be used, or should be used very carefully, among vulnerable women. We suggest that questionnaires and qualitative methods need to include concrete objects, actions, or situations in the process of studying empowerment. This piece helps to initiate a debate about which are the best ways to measure women s empowerment. As researchers who have gone into the field and analyzed the data, we provide evidence about the challenges we found to implementing specific instruments commonly used to capture subjective dimensions of empowerment. We do not mean to suggest that subjective measures of women s empowerment never work. Rather, this evidence points to the fact that it is necessary to contextualize instruments taking into consideration the culture, the education level, the poverty condition and the intersectionalities of women being surveyed. Decontextualized instruments may deeply affect and bias the results. We learned that it is key to take these challenges into account when working with poor women in Latin American countries. We are initiating a study including a large sample of field experiences to measure women s economic empowerment, with a focus on subjective dimensions, which will include more examples from Latin America, Asia, and Africa. We hope that this will be a relevant contribution to developing and using more accurate definitions and measurements of women s economic empowerment. 4. References Arawal, B. (1997). Bargaining and gender relations: within and beyond households. Feminist Economics (3), Alkire, S. (2005) Subjective quantitative studies of human agency, Social Indicators Research, 74 (1), pp Alkire, S., Meinzen-Dick, R., Peterman, A., Quisumbing, A. R., Seymour, G., and Vaz, A. (2013). The Women s Empowerment in Agriculture Index. World Development, 53: Duflo, E. (2012). Women Empowerment and Economic Development. Journal of Economic Literature, Vol. 50, No. 4. Folbre, N. (1994). Who pays for the kids?: Gender and the structures of constraint. London: Routledge. Diener, E., & Biswas-Diener, R. (2002). Will money increase subjective well-being? Social Indicators Research, 57, Ibrahim, S. & Alkire; S. (2007) Agency and Empowerment: A proposal for intentionally comparable Indicator, OPHI Working Paper Series. 9

10 Kabeer, N. (2001). Reflections on the Measurement of Women s Empowerment, Theory and Practice, In Discussing women s empowerment theory (Stockolm, Novum, Grafiska, AB). Kabeer, N. (2005). Gender equality and women's empowerment: a critical analysis of the third millennium development goal 1. Gender & Development, 13(1). Kabeer, N. (2007). Marriage, motherhood and masculinity in the global economy: reconfigurations of personal and economic life. Brighton: Institute of development studies. Kabeer, N., Mahmud, S., & Tasneem, S. (2011). Does paid work provide a pathway to women's empowerment? Empirical findings from Bangladesh. Karlan, D., Ashraf, N., y Yin, W. (2007) Female Empowerment: Impact of a Commitment Savings Product in the Philippines. Center for Global Development, Working Paper No Knowles, J. (2015) Monitoring and evaluation guidelines for women s economic empowerment programs. prepared as a companion to the report, Measuring Women s Economic Empowerment, published by the United Nations Foundation and the ExxonMobil Foundation Molyneux, M. (2008). Conditional Cash Transfers: A 'pathway to women's empowerment'? Brighton: Institute of Development Studies. Mullainathan, S., & Shafir, E. (2013). Scarcity: Why having too little means so much. New York, NY: Times Books. Maslow, A. H. (1970). Motivation and personality. New York: Harper & Row. Narayan, D., Chambers, R., Shah, M.K and Petesch, P. (2000) Voices of the Poor: Crying Out for Change, Washington DC: Oxford University Press for the World Bank. Petesch, P., Smulovitz, C., & Walton, M. (2005). Evaluating empowerment: A framework with cases form Latin America. In D. Narayan, Measuring empowerment: Cross-disciplinary perspectives. Washington D.C.: World Bank. Samman, E.; Santos, E. (2009). Agency and Empowerment: A review of concepts, indicators and empirical evidence. OPHI Working Paper Series. Prepared for the 2009 Human Development Report in Latin America and the Caribbean. Sen, A. (1981). Poverty and famines: an essay on entitlement and deprivation. Delhi: Oxford University Press. Sen, A. (1999). Development as freedom. Oxford: Oxford University Press. Tay, L.; & Diener, E. (2011). Needs and Subjective Wellbeing Around the World. Journal of Personality and Social Psychology. Vol. 101, No2, World Bank (2001) World Bank Development Report: Attacking Poverty (New York, Oxfrod University Press). 10

11 United Nations Foundation and ExxonMobil Foundation (2015). Measuring Women s Economic Empowerment. Washington DC. 11